AIM: A retrospective study of 69 cases of gastric cancer seen during the period from 1/1-1990 to 31/12-1994 treated in a University Hospital. The aim of the study was to describe morbidity, mortality and identify independent prognostic variables for mortality. METHOD: Patient data were recovered from the hospital's central database. Mortality was chosen as end-parameter. Univariate log-rank-test identified statistically significant variables which were then analysed by Cox backward stepwise regressional analysis. MATERIAL: Sixty-nine patients were available for analysis, median age 73 years. Fifty-one patients underwent operation. Eighteen patients did not have a surgical procedure due to disseminated disease. The overall postoperative morbidity was 25% and postoperative mortality 10%. The overall five-year survival rate was 8%, 12% for operated patients, 35% after radical and 0% after non-radical or omitted surgery. Age, radicality of operation, type of operation, Borrmann's tumour classification, and degree of depth of local infiltration were identified as significant factors for survival. Cox's analysis identified type of operation (p = 0.0002) and Borrmann's tumour classification (p = 0.001) as independent variables. DISCUSSION: The overall five-year survival is low and has not changed over two decades in Denmark, whereas mortality and morbidity rates have improved. It should be recommended that: The treatment of gastric cancer must be centralised in order to develop preoperative examinations, operative technique and the necessary routine for the surgeons. All gastric ulcers must be considered malignant and biopsies taken accordingly.
Ugeskrift for Laeger, 2001, Vol 163, Issue 9, p. 1260-4